Introduction
Volatile organic compounds (VOCs) are some types of organic chemicals with high vapor pressure at room temperature.1 VOC substances include alkanes, aromatics, ketones, paraffins, alcohols, esters, ethers and so on.2These hazardous chemicals are emitted from exhaust of industrial plants and involve in atmospheric photochemical reactions. Since these air pollutants cause long-term health problems and environmental issues, control of VOCs emission seems to be a major concern in quality of air. To control VOC emission into atmosphere, many technologies have been proposed such as biodegradation, condensation, catalytic oxidation, adsorption, absorption, etc.3When VOC high removal efficiency and good operation adaptation is concerned, the adsorption process is considered as an effective tool to treat VOCs. 4 In adsorption process, VOCs are held on surface of adsorbent and its pores using the vander Waals force. Activated carbon is widely used as a suitable adsorbent for VOCs recovery due to its large surface area, high adsorption capacity, non-selective nature and low cost comparing to zeolites.5-6However, activated carbon could not act as a catalyst properly and more appropriate adsorbent may be explored. When VOCs are passed over activated carbon, they are adsorbed on the carbon surface and treated air is exhausted to at atmosphere. When all surfaces of activated carbon are occupied, the adsorbent undergoes regeneration to release VOCs by heating it with steam in the temperature swing adsorption (TSA) system. As long as activated carbon gets warmer, it holds less VOCs and then the regeneration stream, a mixture of VOC and steam, exits from the bed and condensed. After cooling stage, the carbon is now ready to be re-used for adsorption.
Shah et al.,7 investigated adsorption of acetone and methyl ethyl ketone (MEK) on activated carbon and its regeneration via hot air in a TSA system. They observed 95 % adsorption capacity for acetone at 80 °C and continuous degradation of the adsorption capacity for MEK. Wang et al.,8 investigated adsorption process for automotive painting components on beaded activated carbon and used hot nitrogen for desorption. They observed competitive adsorption for mixture and displacing of low boiling point compounds with high boiling point compounds. They found that high boiling point compounds may show “heel accumulation” on activated carbon due to their low desorption rate. Tefera et al.,9 studied two-dimensionally model for adsorption of acetone, benzene, toluene and 1,2,4-trimethylbenzene in a fixed-bed cylindrical adsorber over beaded activated carbon. They investigated the effect of operating parameters (namely; temperature, superficial velocity, VOC load, particle size, etc) on the process efficiency and properly simulated transport phenomena in adsorber column. Kim et al., 10 used X- or Y-type Faujasite and Mordenite zeolites to adsorb VOCs and microwave heating for desorption. The highest adsorption capacity was attributed to Faujasite zeolite due to its large surface area and mesoporous volume. They finally concluded that pore structure of zeolites controls adsorption properties of zeolites. Wang et al., 11 synthesized a series of polymeric adsorbents to investigate dichloromethane and 2-butanon adsorption/desorption performance. They found that high surface area enhances adsorption of medium to high concentration of VOCs and meso-pores increase desorption efficiency. Al-Ghouti et al.,12 characterized diethyl ether adsorption using an innovative refrigerator. They described adsorption data by the Langmuir model and chemisorption was attributed as the dominant mechanism for diethyl ether adsorption over activated carbon. An et al.,13 used the Grand Canonical Monte Carlo to simulate adsorption of VOCs on activated carbon. They found that although acetone and methanol can reach to their best adsorption capacity on activated carbon, benzene and toluene functional groups on reduces their adsorption performance. Laskar et al., 14described competitive adsorption of water and multi-component of VOC over activated carbon. They properly fitted experimental data with that of model and proposed the modified Dubinin-Radushkevich and Qi-Hay-Rood models for VOCs and water vapor adsorption, respectively. Gabrus and Downarowicz6 developed two combined TSA systems for ethanol recovery from wet air over molecular sieve 3Aand activated carbon Sorbonorit-4.They utilized the Thomas model to predict breakthrough curve of ethanol and water in liquid and vapor phase.
Up to now, removal of VOCs is studied extensively and most of researchers only focused on adsorption state than cyclic one. In this study, industrial cyclic TSA unit is mathematically modeled to simultaneous removal of diethyl ether and ethanol from air. The breakthrough curves are plotted and compared to industrial data while good agreement is seen between them. With all progresses in modeling, simulation and optimization of TSA systems, there is a still shortcoming in multi-objective optimization. Based on open literature, there is scarce detail study attempting to model a TSA unit satisfying multi-objective optimization. Since, the regeneration step is the most important stage in the TSA unit, four variables relating to the regeneration step are considered as effective parameters. In order to reach the lowest operating costs and highest ethanol and diethyl ether recovery in the TSA unit, multi-objective optimization is conducted as a new study using the response surface methodology (RSM). Eventually, a set of regeneration variables and their optimal values are obtained and a model is proposed for individual objectives.